Integrating SWAT+ hydrological modelling and remote sensing analysis to estimate surface water balance and groundwater recharge in the High Atlas of Marrakech and the Haouz plain (Morocco)

Abstract

Groundwater resources are becoming increasingly vulnerable to human activities and climate change due to high water demand, intensive abstraction, and increased evapotranspiration associated with changes in precipitation amounts and patterns. In semi-arid mountain watersheds, groundwater recharge estimation is affected by substantial uncertainty due to the complexity of hydrological processes and the limitations of individual methods. Process-based models, satellite-derived water balance approaches, and groundwater-level analyses capture different components and scales of recharge, and no single method can fully represent recharge dynamics. To address this, this study applies an integrated framework combining SWAT+ hydrological modelling, remotely sensed water balance components, and the Water Table Fluctuation (WTF) method in the Ourika watershed, originating in the High Atlas of Marrakech and draining into the Haouz plain (Morocco). The dominant land use types of the watershed are grasslands and bare lands. The inputs of the SWAT+ model were prepared using the SRTM 30m digital elevation model (DEM). The improved maps of land use land cover from (ESA CCI LC) products were used to test different scenarios and their impact on the water balance. The soil characteristics are determined from FAO soil maps and the Harmonized World Soil Database and hydraulic characteristics are determined using the SPAW model. Daily rainfall is measured at gauge stations, and the meteorological variables such as daily wind speed, relative humidity, solar radiation, and temperature are collected within the watershed. The model was calibrated using daily stream flow data using Sequential Uncertainty Fitting (SUFI-2), which is one of the programs incorporated into R-SWAT interface. The annual recharge, calculated through the physically-based SWAT+ model, was compared to the estimated one using a remote sensing-based water balance approach. For this latter one, the water balance elements were calculated using satellite-derived datasets from GPM (precipitation), SEBAL (actual evapotranspiration), SMAP (soil moisture storage change), and NRCS-CN (runoff). This method provides spatially distributed estimates of recharge and enables direct comparison with SWAT+ outputs. In addition, the Water Table Fluctuation (WTF) method derived from piezometric time series was used as an independent field-based evaluation of recharge dynamics. The integration of both modeling and remote sensing approaches enhances the understanding of recharge dynamics in semi-arid environments and supports the development of sustainable groundwater management strategies in the Ourika watershed.